Journal of Human Hypertension (2008) 22, 596–607 & 2008 Macmillan Publishers Limited All rights reserved 0950-9240/08 $30.00 www.nature.com/jhh
ORIGINAL ARTICLE
Pharmacological interventions for hypertensive emergencies: a Cochrane systematic review MI Perez and VM Musini Department of Anesthesiology, Pharmacology and Therapeutics, Faculty of Medicine, The University of British Columbia, Vancouver, British Columbia, Canada
Hypertensive emergencies occur when high blood pressure is associated with the presence of acute endorgan damage, such as hypertensive encephalopathy. There is controversy as to when and which antihypertensive drugs to use in these situations. Using a comprehensive search strategy in electronic sources, MEDLINE, EMBASE and Cochrane clinical trial register, we conducted a systematic review to look all randomized control trials (RCTs) that compare an antihypertensive drug versus placebo, no treatment or another antihypertensive drug. Fifteen RCTs (representing 869 patients) met the inclusion criteria. Two trials included a placebo arm. All studies (except one) were open-label trials. Seven drug classes were evaluated in those trials: nitrates (nine trials), angiotensin-converting enzyme inhibitors (seven), diuretics (three), calcium channel blockers (six), a-1 adrenergic antagonists (four), direct vasodilators (two) and dopamine agonists (one). Mor-
tality event data were reported in seven trials. Due to insufficient data, no meta-analysis was performed for clinical outcomes. Differences in blood pressure changes between antihypertensives were minor. There is no RCT evidence demonstrating that antihypertensive drugs reduce mortality or morbidity in patients with hypertensive emergencies. Furthermore, there is insufficient RCT evidence to determine which drug or drug class is most effective in reducing mortality and morbidity. There were some minor differences in the degree of blood pressure lowering when one class of antihypertensive drug is compared to another. However, the clinical significance is unknown. RCTs are needed to assess different drug classes to determine initial and longer-term mortality and morbidity outcomes. Journal of Human Hypertension (2008) 22, 596–607; doi:10.1038/jhh.2008.25; published online 17 April 2008
Keywords: hypertensive emergencies; randomized controlled trials; systematic review; meta-analysis
Introduction A hypertensive emergency is the clinical setting where a marked elevation of blood pressure is associated with acute end-organ damage, for example, hypertensive encephalopathy or aortic dissection. As such it is a life-threatening condition. The goal of treatment is to reverse the end-organ damage, prevent adverse outcomes and prolong life. This review focuses on blood pressure lowering drugs that are used in this emergency setting. The management of hypertension in these emergency situations represents a significant therapeutic challenge. Many antihypertensive drug classes have been used with the objective of rapidly reducing blood pressure, and the expectation of reducing adverse clinical outcomes. This approach was first Correspondence: Dr MI Perez, Department of Anesthesiology, Pharmacology and Therapeutics, Faculty of Medicine, The University of British Columbia, 2176 Health Science Mall, Vancouver, British Columbia, Canada V6T 1Z3. E-mail:
[email protected] Received 7 December 2007; revised 26 January 2008; accepted 23 February 2008; published online 17 April 2008
recommended by Gifford1 based on a series of eight cases with hypertensive encephalopathy that were treated with sodium nitroprusside. On the basis of this case series evidence this approach has become and remained the standard of care and is currently recommended by most if not all guideline committees (such as JNC-72). At issue in this review is whether randomized control trial (RCT) evidence supports this approach and which drug classes are the most effective. Two published systematic reviews have addressed these issues. One compares different antihypertensive drugs, but it pools hypertensive emergency and urgency trials.3 Urgencies are defined as marked elevated blood pressure in an otherwise stable patient (that is, without acute end-organ damage). In our opinion the urgency setting is very different from that of emergencies and needs to be reviewed separately. The second systematic review, a Cochrane review of interventions that alter blood pressure after acute stroke is not limited to RCTs studying drugs to reduce blood pressure and includes RCTs whether or not the patients had elevated blood pressure.4
Different interventions for hypertensive emergencies MI Perez and VM Musini 597
Therefore, it also does not answer the question raised here. Objectives
General. The general objective is to find and quantify the RCT evidence for antihypertensive drug treatment of patients with a hypertensive emergency, defined as marked hypertension associated with acute end-organ damage. Specific. The specific objective is to answer the following two questions: Does antihypertensive drug therapy as compared to placebo or no treatment affect mortality and morbidity in patients with a hypertensive emergency? Does one first-line antihypertensive drug class offer a therapeutic advantage, in terms of mortality and morbidity, over another in patients with a hypertensive emergency? Criteria for considering studies for this review
Types of studies. All unconfounded, truly RCTs that compare a first-line antihypertensive drug class versus placebo, no treatment or another first-line antihypertensive drug class. Crossover trials were excluded. There was no language restriction. Types of participants. Participants must meet the following hypertensive emergency definition: any clinical setting where patients present with marked elevation of blood pressure in the presence of acute end-organ damage. Examples of acute end-organ damage are the following: myocardial infarction, unstable angina, acute left ventricular failure with pulmonary oedema, acute aortic dissection, encephalopathy, stroke and life-threatening bleeding (intracerebral haemorrhage and subarachnoid haemorrhage). Thus, patients with marked elevation of blood pressure but without acute end-organ damage (defined as urgencies) were not included. There is no evidence as to what constitutes ‘marked blood pressure elevation’. Therefore, we have chosen blood pressure level(s) commonly used in clinical practice to mandate the use of antihypertensive drugs (along with other acute therapy such as pain management) in relevant clinical settings. For example, for patients with acute myocardial infarction a systolic blood pressure (SBP) greater or equal to 180 and or diastolic blood pressure (DBP)X110 mm Hg is the threshold above which thrombolysis is contraindicated (ACC/ AHA-5). For patients with acute aortic dissection or with left ventricular failure and pulmonary oedema an SBP greater or equal to 120 mm Hg and or DBPX70 mm Hg is the threshold for therapy.6,7 For patients with intracranial haemorrhage or subarachnoid haemorrhage an SBPX160 mm Hg is the
threshold because of a higher incidence of rebleeding above this level.8 For patients with any other acute end-organ damage setting an SBPX180 and or DBPX110 mm Hg is the defined threshold. We included all RCTs that included patients with these minimum or higher thresholds. In the case that an RCT does not define blood pressure inclusion criteria but had included only one category of patients (patients with pulmonary oedema, for example), then the mean baseline blood pressure had to be equal to or greater than these defined thresholds. In the event that an RCT had included patients with different end-organ damage clinical settings, a mean baseline blood pressure of SBPX180 and or DBPX110 mm Hg is acceptable for inclusion. Note that pregnancy-related hypertensive emergencies are excluded from this review. Types of interventions
Intervention: A first-line antihypertensive drug class (first-line antihypertensive drug classes include: nitrates, b-blockers, angiotensin-converting enzyme (ACE) inhibitors, diuretics, calcium channel blockers (CCBs), dopamine agonists, a-adrenergic antagonists and direct vasodilators (diazoxide and hydralazine). Control: placebo, no treatment or a different firstline antihypertensive drug class. Types of outcome measures
Primary outcome measures include: total serious adverse events, all-cause mortality and composite of non-fatal cardiovascular events, including myocardial infarction, unstable angina, dissection of aortic aneurysm, acute renal failure, stroke and respiratory failure (necessitating mechanical ventilation). Secondary outcome measures include: weighted mean change in SBP, DBP and in heart rate (HR), during the treatment period and withdrawals due to adverse effects.
Methods of the review Data abstraction
The search strategy for identification of studies is summarized in Table 1. Two reviewers (MIP and MVM) independently decided whether a trial was included. They also independently extracted and entered the data from the included studies. Discrepancies were resolved by discussion. Absence of consensus was resolved by a third reviewer (JMW). A modified Cochrane quality scoring system was used for concealment of allocation and blinding: A Journal of Human Hypertension
Different interventions for hypertensive emergencies MI Perez and VM Musini 598
Table 1 Cochrane search strategy: see also Collaborative review group search strategy We searched randomized controlled trials of all antihypertensive drugs used for hypertensive emergencies through the following database of articles published from 1966 to August 2007, MEDLINE, EMBASE and Cochrane clinical trial register. A comprehensive search strategy was used to identify all relevant articles. Review articles and trials reference lists were also checked. Keywords: controlled clinical trial, randomized controlled trials, meta-analysis, severe/accelerated/crisis(es), hypertension, antihypertensive, emergencies: hypertensive encephalopathy, myocardial infarction, unstable angina, acute left ventricular failure, pulmonary oedema, stroke, subarachnoid/ intracranial haemorrhage, aortic dissection and list of drugs as follows. 1 Randomized controlled trial.pt. 2 Randomized controlled trials.mp. 3 Randomized controlled trial.mp. 4 Controlled clinical trial.pt. 5 Controlled clinical trials.mp. 6 Controlled clinical trial.mp. 7 Random allocation.mp. 8 Exp double-blind method/ 9 Double-blind.mp. 10 Exp single-blind method/ 11 Single-blind.mp. 12 or/1–11 13 Exp animal/ 14 12 not 13 15 Clinical trial.pt. 16 Clinical trials.mp. 17 Clinical trial.mp. 18 Exp clinical trials/ 19 (Clin$ adj25 trial$).mp. 20 ((singl$ or doubl$ or trebl$ or tripl$)adj25 (blind$ or mask$)).mp. 21 Random$.mp. 22 Exp research design/ 23 Research design.mp. 24 or/15–23 25 24 not 13 26 25 not 14 27 Comparative studies.mp. 28 Comparative study.mp. 29 Exp evaluation studies/ 30 Evaluation studies.mp. 31 Evaluation study.mp. 32 Follow up studies.mp. 33 Follow up study.mp. 34 Prospective studies.mp. 35 Prospective study.mp. 36 (control$ or prospective$ or volunteer$).mp. 37 or/27–36 38 37 not 13 39 38 not (14 or 26) 40 14 or 26 or 39 41 Alacepril.mp. 42 Benazepril.mp. 43 Captopril.mp. 44 Ceronapril.mp. 45 Cilazapril.mp. 46 Derapril.mp. 47 Enalapril.mp. 48 Enalaprilat.mp. 49 Fosinopril.mp. 50 Idapril.mp. 51 Imidapril.mp. 52 Lisinopril.mp. 53 Moexipril.mp. 54 Moveltopril.mp. 55 Perindopril.mp. 56 Quinapril.mp. Journal of Human Hypertension
Table 1 Continued 57 58 59 60 61 62
Ramipril.mp. Spirapril.mp. Temocapril.mp. Trandolapril.mp. Zofenopril.mp. Angiotensin converting enzyme inhibitor.mp. or angiotensinconverting enzyme inhibitors/ 63 Acebutolol.mp. 64 Atenolol.mp. 65 Bisoprolol.mp. 66 Esmolol.mp. 67 Labetalol.mp. 68 Metoprolol.mp. 69 Nadolol.mp. 70 Practolol.mp. 71 Propranolol.mp. 72 Sotalol.mp. 73 Timolol.mp. 74 Carvedilol.mp. 75 Adrenergic beta-Antagonists.mp. 76 Amlodipine.mp. 77 Aranidipine.mp. 78 Azelnidipine.mp. 79 Barnidipine.mp. 80 Bencyclane.mp. 81 Benidipine.mp. 82 Bepridil.mp. 83 Cilnidipine.mp. 84 Cinnarizine.mp. 85 Clentiazem.mp. 86 Darodipine.mp. 87 Diltiazem.mp. 88 Efonidipine.mp. 89 Elgodipine.mp. 90 Etafenone.mp. 91 Fantofarone.mp. 92 Felodipine.mp. 93 Fendiline.mp. 94 Flunarizine.mp. 95 Gallopamil.mp. 96 Isradipine.mp. 97 Lacidipine.mp. 98 Lercanidipine.mp. 99 Lidoflazine.mp. 100 Lomerizine.mp. 101 Manidipine.mp. 102 Mibefradil.mp. 103 Nicardipine.mp. 104 Nifedipine.mp. 105 Niguldipine.mp. 106 Nilvadipine.mp. 107 Nimodipine.mp. 108 Nisoldipine.mp. 109 Nitrendipine.mp. 110 Perhexiline.mp. 111 Prenylamine.mp. 112 Semotiadil.mp. 113 Terodiline.mp. 114 Tiapamil.mp. 115 Verapamil.mp. 116 Calcium channel blocker.mp. or calcium channel blockers/ 117 Nitroprusside.mp. 118 Nitroglycerine.mp. 119 Nitroglycerin/ or nitroglycerine.mp. or isosorbide dinitrate/ 120 Nitrates.mp. or nitrates/ 121 Urapidil.mp. 122 Trimethaphan/ or trimethaphan camsylate.mp. 123 Reserpine.mp. 124 Phentolamine.mp. 125 Methyldopa.mp. 126 Labetalol.mp. 127 Ketanserine.mp.
Different interventions for hypertensive emergencies MI Perez and VM Musini 599
Table 1 Continued 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
Hydralazine.mp. Guanethidine.mp. Fenoldopam.mp. Diazoxide.mp. Clonidine.mp. Thiazide$.mp. Hydrochlorothiazide.mp. Chlorthalidone.mp. or chlorthalidone/ Furosemide.mp. or furosemide/ or/41–136 40 and 137 Myocardial infarction.mp. Unstable angina.mp. Acute left ventricular failure.mp. Pulmonary edema/ or pulmonary oedema.mp. Stroke.mp. Life-threatening bleeding.mp. Aneurysm, dissecting/ or aortic dissection.mp. Intracranial hemorrhages/ or cerebral hemorrhage/ or intracranial haemorrhage.mp. Intracranial aneurysm/ or subarachnoid hemorrhage/ or subarachnoid haemorrhage.mp. or/139–147 Hypertension.ti,ab. High blood pressure.ti,ab. Blood pressure.ti,ab. or/149–151 Pulmonary artery hypertension.mp. Pulmonary hypertension.mp. Portal hypertension.mp. or/153–155 152 not 156 148 and 157 Hypertensive emergencies.ti,ab. Hypertensive emergency.ti,ab. Hypertensive urgency.ab,ti. Hypertensive urgencies.ti,ab. Hypertensive crisis.ti,ab. Hypertensive crises.ti,ab. Acute end-organ damage.mp. or/158–165 138 and 166
combined using a weighted mean difference (WMD) method, whereby the trials are weighted according to the number of subjects in the trial and the within-study variance. Some of the trials did not report a within-study variance for blood pressure reduction. In these studies standard deviation (s.d.) was imputed using the following hierarchy: (1) Pooled s.d. calculated either from the t-statistic corresponding to an exact p-value reported or from the 95% CI of the mean difference between treatment group and comparative group. (2) Standard deviation of blood pressure/HR at the end of treatment. (3) Standard deviation of blood pressure/HR at baseline (except if this measure is used for entry criteria) (4) Weighted mean s.d. of change in blood pressure/ HR calculated from at least (5) Other trials using the same drug and dose regimen (6) Weighted mean s.d. of change in blood pressure/ HR calculated from other trials using the same drug (7) Weighted mean s.d. of change in blood pressure/ HR calculated from all other trials (any drug and dose) Several sensitivity analyses were pre-planned to test robustness, including the use of both fixed and random effects models, 95 and 99% CIs, and quality of trials. Also sensitivity analyses were pre-planned according to the clinical setting and to the class of drug.
Results Description of studies
(adequate and double blind), B (unclear and single blind or open label) and C (clearly inadequate and open label). The two reviewers (MIP and MVM) also independently assessed the quality of studies. Authors were contacted in case of missing information. Analyses
For the synthesis and analysis of the data Cochrane review manager software Revman 4.2.9 was used. Relative and absolute risk differences (with 95% confidence interval (CI)) were calculated for dichotomous outcomes for each trial on an intention to treat basis. Heterogeneity between trial results was tested using w2-test, where po0.05 was taken to indicate significant heterogeneity. The fixed effect model was used when there was homogeneity and the random effect model was used to test for statistical significance where there was heterogeneity. Trials were not subclassified according to dose or dosing regimen. Data for blood pressure was
Our search strategy yielded 86% of irrelevant studies in the first screening stage by reading titles and abstracts (Figure 1). Fifteen RCTs (869 patients) were found that satisfied the inclusion criteria9–23 (Table 2). Two trials were placebo controlled.12,18 Only one trial12 was confirmed to be double blind, whereas the rest were open label. No trial was designed for or had the power to detect differences in clinical outcomes .The largest trial consisted of 133 patients.20 The longest trial11 lasted 10 days. Most of the trials reported data for only 2–6 h. Seven drug classes were evaluated: nitrates (9 trials), ACE inhibitors (7), CCBs (6), peripheral a-1 blockers (4), diuretics (3), direct vasodilators (2) and dopamine agonists (1). All included trials had patients with elevated blood pressure in the presence of acute end-organ damage. Blood pressure entry criteria differed among trials. Four trials were included on the basis of their mean blood pressure values at baseline.10,12,17,18 Seven trials included exclusively patients with acute pulmonary oedema.10,12,14,17,20,21,23 One trial included exclusively Journal of Human Hypertension
Different interventions for hypertensive emergencies MI Perez and VM Musini 600
Citations identified in literature search: N=5,413 Citations excluded by reading title and abstract. (clearly no relation to our work) n=4,660
Citations retrieved for more detailed evaluation: n=753 Excluded: n=711 Reasons : Non-randomized: 354 Non-acute end organ damage (AEOD) setting: n=270 (AEOD) setting but BP below our pre-established thresholds defining a hypertensive emergency: n=87
Potentially appropriate randomized controlled trials: n=42
Excluded: n=27 Reasons : (see text for details)
Included: n=15 Figure 1 Quorum flowchart of studies.
patients with hypertensive encephalopathy.16 There was no trial that included exclusively patients with acute aortic dissection or acute myocardial infarction. Thus, the rest of seven trials included a diverse population with different acute end-organ damage. Only two trials9,15 reported the s.d. of the change of blood pressure. In the rest of the trials this measure of variability was imputed from the s.d. at end point. Additional information was required and requested from all included trials. One trialist9 provided missing information in the original publication. The rest of the trialists did not reply to our request. We excluded 27 clinical trials for several reasons: Several trials mixed patients with and without acute end-organ damage in the same RCT (12 trials).24–35 Other trials included patients without explicitly stating whether patients had acute end-organ damage or not (7 trials).36–42 Some trials included non-randomized participants in the trial’s results (1 trial).43 One trial did not report any of the outcomes of interest (1 trial).44 Two trials did not fulfil blood pressure threshold criteria.45,46 Journal of Human Hypertension
One was a cross-over trial.47 Two trials had wrong comparators (1 compared different doses of the same combination therapy;48 1 compared two drugs of the same class.49 One RCT only included responders to a previously given antihypertensive therapy.50 Two out of 27 excluded trials involved a b-blocker arm and 18/27 excluded trials involved a calcium channel blocker arm. One excluded trial studied exclusively patients with acute aortic dissection.49 Methodological quality of included studies
All studies, except one,12 were open-label trials. The method of randomization was not reported in eight trials. The method to achieve concealment of allocation was reported in only two trials.12,16 Comparisons according to outcomes
Total serious adverse events. No trial reported total serious adverse events. All-cause mortality. Mortality was reported in seven trials9,10,14,16,17,20,21 and totalled six deaths in three RCTs. The group to which the dead patients were originally allocated was not reported for five of the
Table 2 Summary of included studies Author
Angeli et al.9 Beltrame et al.
10
Elliot et al.11
Halminton et al.12 13
Hirschl et al.
Hirschl et al.14 Marigliano et al.15 Mcnair et al.16 Nelson et al.17
Rubio et al.19 Schreiber et al.20 Verma et al.21 Wu et al.22 Yang et al.23
n
Blood pressure (mm Hg) inclusion criteria
Mean SBP/DBP (mm Hg) at baseline
Clinical inclusion criteria
Mean SBP/DBP (mm Hg) at end point
Mortality
Nifedipine 10 mg s.l. Captopril 25 mg (s.l.) Nitroglycerine 2.5–10 mg i.v. infusion Furosemide 40 mg i.v. boluses Nitroprusside 0.5-mg kg1 min1 i.v. infusion Fenoldopan 0.1 mg kg1 min1 i.v. infusion Placebo Captopril 25 mg (s.l.) Nitroprusside 0.5–3 mg k min1 i.v. infusion Urapidil 12.5 mg i.v. boluses Nitroglycerin 0.8 mg (s.l.) Enalaprill 2.5 mg i.v. boluses Nifedipine 10 mg (s.l.) Captoprill 25 mg (s.l.) Diazoxide 75–150 mg i.v. boluses Dihydralazine 6.25–12.5 mg i.m. boluses Isosorbide 50–200 mg kg1 h1 i.v. infusion Furosemide 1 mg kg1 i.v. boluses Placebo Nifedipine 10 mg (s.l.) Captopril 50 mg (s.l.) and oral Ketanserine 40 mg (s.l.) Isosorbide aeroso 1.25 mg oral Nifedipine 10 mg (s.l.) Nitroglycerine 0.8 mg (s.l.) Urapidiil 12.5 mg i.v. boluses Furosemide 1 mg kg1 i.v. bolus Isosorbide 50–200 mg i.v. infusion Hydralazine 0.15 mg kg1 i.v. bolus Nifedipine 10 mg (s.l.) Captopril 25 mg (s.l.) Prazosin 10 mg (s.l.) Nitroprusside 1 mg kg1min1 i.v. infusion Nicardipine 3 mg kg1min1 i.v. infusion
10 10 37 32 15 13
DBP4140
247/158 245/145 161/NR 164/NR 222/137 214/136
Acute end-organ damagea Acute pulmonary oedema Acute end-organ damagea
204/115 190/116 133/NR 139/NR 174/105 180/106
0 0 3
160/100 172/112 211/109 215/107
Acute pulmonary oedema Acute end-organ damagea
NR NR 151/74 162/88
Not reported
206/116 211/115 208/139 230/120 228/118 218/142 130/75 119/72
Acute pulmonary oedema Acute end-organ damagea Hypertensive encephalopathy Acute left ventricular failure
136/71 139/70 154/70 163/90 179/110 169/101 118/70 116/70
188/111 195/114 196/114 185/117 187/121 190/115 216/116 218/118 117/73 131/75 134/77 198/124 198/122 198/128 195/115 196/114
179/101 162/92 168/97 165/100 Acute end-organ 153/92 a damage 153/86 Acute pulmonary 134/72 oedema 134/70 Acute left ventricular 112/70 failure 118/70 128/71 Cerebral signs or 140/78 symptoms 134/78 160/90 Acute pulmonary 144/85 oedema 135/79
25 23 35 46 23 23 22 22 28 24 14 14 20 16 41 15 30 30 73 60 12 12 12 30 35 27 20 20
Not defined
b
DBP4120
Not definedb SBP4200 and/or DBP4110 SBP4200 or DBP4100 SBP4210 DBP4135 Not definedb Not definedb
MAP4130 SBP4200 DBP4100 SBP4100 SBP4190 and/or DBP4120 SBP4160 and/or DBP4100
Acute end-organ damagea
Not reported
Not reported 0 Not reported 2 0 0 Different interventions for hypertensive emergencies MI Perez and VM Musini
Pastorelli et al.18
Comparators and dose
Not reported
Not reported 0 0 0 0 1 Not reported Not reported
601
Journal of Human Hypertension
Abbreviations: DBP, diastolic blood pressure; i.m., intramuscular; i.v., intravenous; MAP, mean arterial pressure; SBP, systolic blood pressure; s.l., sublingual. a As stated in the article reflecting the inclusion of patients with different acute end-organ damage settings. b This RCT was included on the basis of the mean blood pressure values at baseline according to our pre-defined thresholds for this category of patients.
Different interventions for hypertensive emergencies MI Perez and VM Musini 602
deaths. In one RCT, a patient treated with hydralazine died of a rupture of the inter-ventricular septum.21 In four trials mortality was reported as nil. In eight trials there was no mention of mortality. It is possible that there were no deaths during the short range of followup (6–24 h), but it is impossible to be certain. Non-fatal cardiovascular events Composite. Cardiovascular events were reported in five trials.10,12,14,16,20 No trial reported cardiovascular events as a composite. It was not possible to extract events from the original trials and analyse them as a composite due to a risk of double-counting the events. Myocardial infarction. One placebo-controlled trial12 reported this outcome. There was no statistically significant difference between ACEi and placebo (RR 0.72, 95% CI 0.31–1.72). Three head-to-head trials reported this outcome.10,16,20 There was no statistical difference in myocardial infarction between nitrates (2.7%) and aadrenergic antagonist (5%) (RR 0.55, 95% CI 0.09– 3.17); or nitrates (16%) vs diuretics (12.5%) (RR 1.30,95% CI 0.40–4.19); or between diazoxide (3.5%) vs dihydralazine (4%), (RR 0.86, 95% CI 0.06–12.98). Pulmonary oedema requiring mechanical ventilation. Three trials reported this outcome.12,14,20 There was no meta-analysis performed since there was only one trial for each comparison. There was no statistically significant difference between captopril and placebo (RR 0.40, 95% CI 0.09–1.86), nitrates and a-adrenergic antagonist (RR 4.12, 95% CI 0.20–84.24 or between nitrates and ACE Inhibitor (RR 0.33, 95% CI 0.01–7.78). Other than the above, the trials did not report any of our list of cardiovascular events (unstable angina, dissection of aortic aneurysm, acute renal failure or stroke). An additional cardiovascular event was reported that was not on our list: asystole, which happened in one patient randomized to an ACE inhibitor.14 Withdrawals due to adverse events. Only one trial comparing an a-blocker with nitroglycerine reported withdrawal due to adverse events.20 There were no significant differences between these two drugs classes (5 vs 2.7%; RR 3.38, 95% CI 0.17–68.84). Weighted mean change in blood pressure and HR during treatment. For this secondary outcome all trials provided some data and we were able to pool this data. Drug vs placebo or no treatment. Although we included two placebo-controlled trials, only one provided SBP or DBP data18 and this was limited to 1 h of follow-up. In this trial, three classes of antihypertensives were included: calcium channel blocker, ACE inhibitors and a-1 adrenergic antagonists (A1A). The pooled effect showed a statistically Journal of Human Hypertension
significant greater reduction in both, SBP (WMD 13.14, 95% CI, 19.48,6.80) and DBP (WMD 8.03, 95% CI, 12.61,3.45) with antihypertensive therapy. There were no data on HR. It was not possible to extract blood pressure data from the other placebo-controlled trial.12] In addition to not reporting any measurement of variability, this trial reported blood pressure data as change in mean arterial pressure (MAP). Nitrates vs diuretics. Three trials compared nitrates to diuretics (Beltrame, Nelson and Verma)10,17,21. Furosemide was the common diuretic used in all of them with two nitrates, nitroglycerine and isosorbide as comparators. Neither SBP- nor DBP lowering effect was statistically different between the two classes of drugs. However, in Beltrame,10 the SBP lowering effect of both drugs was greater (21 mm Hg for furosemide; 23.75 mm Hg for nitroglycerin) than that reported in the other two trials ( þ 1.0, þ 1.6 mm Hg for furosemide groups and 6, 8 mm Hg for isosorbide groups, respectively). The reasons for that difference across trials are not clear. Despite these differences, heterogeneity was not present when pooling all these three trials. HR change was also not significantly different for both classes of drugs. Nitrates vs a-1 antagonist. Two trials compared the A1A, urapidil, with nitrates.13,20 The first trial used nitroprusside and the second used nitroglycerine as comparator. The SBP lowering effect of the two nitrates was similar (58.4 mm Hg for nitroprusside and 59.5 mm Hg for nitroglycerine). However, the effect of urapidil (administrated at the same dose in both trials) was very different (37.6 and 73.5 mm Hg). A similar discrepancy was seen for DBP. This heterogeneity precluded the pooling of these trials in a meta-analysis for these outcomes. Nitrates vs dopamine agonist. For this comparison one trial was included.11 During 4 h of treatment, nitrates were associated with a statistically significant greater reduction in SBP as compared with a dopamine agonist (WMD 14.00, 95% CI (27.72, 0.28). There were no differences between these classes in DBP or HR. Nitrates vs ACE inhibitors. One trial compared a nitrate with an ACE inhibitor.14 No statistically significant difference was found between the two groups in SBP or DBP or HR. Nitrates vs calcium channel blockers. By pooling two trials19,23 CCBs were not associated with statistically significant differences in SBP or DBP as compared to nitrates. Using the fixed effect model, CCBs were associated with statistically significant increase in HR as compared to the nitrates (WMD 11.76 95% CI (4.45,19.07). However there was significant heterogeneity across trials and this
Different interventions for hypertensive emergencies MI Perez and VM Musini 603
increase was no longer statistically significant when a random effect model was used. Nitrates vs direct vasodilator. For this comparison one trial was included.21 There was no statistical difference in SBP or DBP reduction between the two drugs. There was also no significant difference between these classes in HR change. ACE inhibitors vs calcium channel blockers. Four trials,9,15,18,22 compared an ACE Inhibitor with a CCB. The pooled data show that CCBs were associated with a significantly greater reduction in DBP as compared with ACE-I (WMD 7.86, 95% CI (4.92, 10.81). No statistically significant difference was found between the two groups in the reduction of SBP. In three trials that reported HR changes,9,15,22 CCBs were associated with a significant increase in HR as compared with ACE inhibitors (WMD 22.91, 95% CI (19.8, 26.01). However there was significant heterogeneity across trials and this increase was no longer significant when a random effect model was used. ACE inhibitors vs a-1 adrenergic antagonist. Two trials18,22 compared an ACE Inhibitor with an A1A. Both trials used captopril as comparator but one trial used prazosin and the other used ketanserin. The pooled data show that ACE-I were associated with a significantly greater reduction in both SBP and DBP as compared with A1A (SBP WMD 20, 95% CI (22.85,17.39); DBP WMD 3.70, 95% CI (7.08,0.31). For SBP outcome there was statistically significant heterogeneity across trials. However the difference was still significant when the random effects model was used. No statistically significant difference was found between the two groups in the HR change in the only trial reporting that outcome.22 Diazoxide vs hydralazine. For this comparison one trial,16 which dealt with exclusively hypertensive encephalopathy patients, was included. During 4 h of treatment, hydralazine was associated with a statistically significant greater reduction in both SBP (WMD 13.56, 95% CI (3.06,24.06)) and DBP (WMD 14.67, 95% CI (8.01,21.33)) as compared to diazoxide (WMD 14.00, 95% CI (27.72, 0.28). It is important to mention, though, that there was no measure of variability reported in this trial. Therefore, we imputed the s.d. of the change according to our hierarchy from other trials (last option: weighted mean s.d. of change from all trials; any drug any dose). There were no HR data reported.
Discussion This is the first systematic review investigating mortality and morbidity outcomes for all RCTs of
drug treatment for hypertensive emergencies. A systematic review that combined hypertensive emergencies and urgencies3 did not include 11 trials included in our systematic review. Furthermore, Cherney’s review mixed randomized with nonrandomized trials. The only other relevant systematic review in relation to hypertensive emergencies is that conducted for acute stroke by BASC.4 We excluded one trial46 (n ¼ 16 patients) that the BASC 2001 systematic review had included. The reason for excluding it was because the blood pressure criteria in this trial (4170/95 mm Hg) did not meet our blood pressure threshold criteria (SBPX180 and or DBPX 110 mm Hg). This exclusion does not affect our conclusion for clinical outcomes as this trial did not report clinical outcomes. The other BASC 2001 trials were not included because blood pressure at baseline was not elevated. Thus, these clinical trials did not include hypertensive emergency patients, as we have defined it. One of the limitations in our review is that most of the included trials were small (average 58 patients per trial). Furthermore, with the exception of Hamilton et al.12 all trials were of poor quality. Three included trials deserve further discussion. Hamilton et al.,12 the only double-blind trial, includes patients with acute pulmonary oedema and high blood pressure, and it compared captopril vs placebo. It demonstrates that this high-quality and double-blind trial was ethical and feasible. The Danish II 198616 trial was the only trial that included patients exclusively with hypertensive encephalopathy. This was a well-organized multicentre trial, conducted in Denmark, comparing diazoxide vs dihydralazine. Due to its study design, the ethical committee accepted that the informed consent could not be obtained from patients as all of them had symptoms of hypertensive encephalopathy. A downside of this study is the fact that the trialists reported their results in duplicate publications that did not cite the other publications (the original publication, Krogsgaard et al.,51 is not cited in the other duplicate publications,16,52,53 In addition, blood pressure values were not the same in the different publications, and none of the publications reported measures of SBP or DBP variability. The largest trial, Schreiber et al.,20 included 133 patients with acute pulmonary oedema plus high blood pressure, in an out-of-hospital setting, who were randomized to receive either nitroglycerin or urapidil. The ethical committee (Vienna, Austria) agreed that no informed consent had to be obtained at the time of inclusion for randomization. However, the pitfall of this trial is that 16% of all randomized patients were excluded from the analyses that potentially biased the results. Consistent with this, there was significant heterogeneity when this trial was combined with another trial studying the same comparison groups. Journal of Human Hypertension
Different interventions for hypertensive emergencies MI Perez and VM Musini 604
In 19 of the excluded trials24–42 it was not possible to determine how many patients had acute endorgan damage or merely had elevation of blood pressure. We believe that it would be misleading to include these trials in this review as the impact of antihypertensive drugs is potentially different. If individual patient data could be obtained, the patients with acute end-organ damage could be added to our review. It was perhaps surprising and definitely disappointing that we could find no RCT evidence to answer the first question we have posed: Does antihypertensive therapy as compared to placebo or no treatment change mortality and morbidity in patients with hypertensive emergencies? The one available placebo-controlled trial demonstrated that blood pressure was reduced with drugs as compared to the control treatment, however, it was too small and of too short duration to assess morbidity and mortality. We feel it is important for physicians to know that this is one of the clinical settings where treatment is not supported by RCT evidence. Despite the lack of evidence it is not hard to accept the necessity of lowering blood pressure in those clinical settings where the excessive increases in blood pressure are the cause of the end-organ damage. However, this is not necessarily the best approach in settings where the excessive elevations of blood pressure are probably caused by end-organ damage such as high blood pressure in the presence of a cerebrovascular accident. The presently accepted approach for the immediate treatment of hypertensive emergencies in clinical practice is primarily based on a series of cases published in 1959.1 In this study carried out over a period of 18 months the author demonstrated the ability to reduced blood pressure with nitroprusside, within minutes, in eight patients with hypertensive emergencies (mostly patients with hypertensive encephalopathy), whose blood pressures had remained elevated after treatment with reserpine or hydralazine. However, he did not report clinical outcomes so we do not know whether these patients did better as a result of the blood pressure lowering. Gifford recommended prompt blood pressure reduction in clinical settings other than hypertensive encephalopathy such as intracerebral or subarachnoid haemorrhage or acute left ventricular failure. The lack of RCT evidence leaves the distinct possibility that in some clinical settings defined as hypertensive emergencies immediate antihypertensive therapy could be doing more harm than good. There is a hypertensive emergency not included in the present systematic review, eclampsia. Due to its pathophysiology and the involvement of the infant as well as the mother, we felt this clinical entity must be studied separately from other hypertensive emergencies and include outcomes in the infant as well as the mother. There is a Cochrane systematic review54 that has studied the drugs for treatment of very high blood pressure during Journal of Human Hypertension
pregnancy. However, Duley’s systematic review was not limited to patients with eclampsia and did not separately report outcomes in the eclampsia patients. To the best of our knowledge there is no systematic review dealing exclusively with eclampsia and antihypertensive treatment. Thus, a systematic review in this specific area is currently needed. The present review also does not provide any mortality and morbidity evidence from RCTs to inform clinicians as to which first-line antihypertensive drug class provides more benefit than harm in hypertensive emergencies. This lack of evidence was due to the fact that the trials were too small, did not follow the patients for a long enough period of time and frequently failed to report all important outcomes. In addition all the RCTs except one were open-label trials and therefore concealment of allocation was not possible in most cases. Although, these shortcomings of the trials would not likely affect mortality and morbidity outcomes, they could bias blood pressure and HR data. Neither did we find RCTs that compared different strategies to reduce blood pressure. Thus, how fast or how much blood pressure should be lowered in hypertensive emergencies remains unknown. Although it is unproven, it is highly likely that antihypertensive therapy is an overall benefit in a hypertensive emergency and therefore a placebocontrolled trial to prove this would be unethical. What is clear is that this is a clinical area where properly conducted randomized trials are badly needed. At the present time RCTs could be conducted to compare different drug classes and treatment strategies, for example, aggressive rapid lowering of blood pressure to a target versus lowering the blood pressure slowly at a defined rate such as 5–10% every 2 h. What is also clear from this review is that any trial must follow patients long term and document mortality and morbidity. One of the best examples of an adequate RCT in an emergency setting is the corticosteroid randomization after significant head injury (CRASH) trial,55 where 10 000 patients with acute head injury were randomized to intravenous steroids or placebo. Its approach to handle ethical issues could serve as model when conducting a trial with hypertensive emergency patients.
Conclusions Implications for practice
There is no evidence from RCTs that antihypertensive drugs reduce mortality or morbidity in patients with hypertensive emergencies, defined as marked hypertension associated with acute end-organ damage. Furthermore, there is insufficient RCT evidence to determine which drug or drug class is most effective in reducing mortality and morbidity. There were some minor differences in degree of blood
Different interventions for hypertensive emergencies MI Perez and VM Musini 605
pressure lowering between drug classes. However, the clinical significance is unknown. This review demonstrates a blood pressure lowering efficacy for nitrates, ACE inhibitors, diuretics, a-adrenergic antagonist, CCBs and dopamine agonists. Nitrates (including nitroprusside) have been studied in most. Therefore, if a hypertensive emergency patient cannot be treated as part of an RCT and a nitrate is available, it is a reasonable choice of therapy.
6 7 8
Implications for research
RCTs are needed to assess different blood pressure lowering strategies and different first-line drug classes in patients with hypertensive emergencies. Outcomes in such trials must be mortality and total serious adverse events at different times of followup such as 7 days, 1 month and including at least 6 months of follow-up of all patients.
Acknowledgements We acknowledge help and advice from Dr Ken Bassett, Stephen Adams for retrieving trials, and Benji Heran, Michelle Wong and Jenny Chang for comments on a draft. We also acknowledge the trialists who provided us with additional information from their studies. This paper has also been published in the Cochrane Library as: MI Perez, VM Musini. Pharmacological interventions for hypertensive emergencies. Cochrane Database of Systematic Reviews 2008, Issue 1. art no.: CD003653. doi: 10.1002/14651858.CD003653.pub3.
9
10
11
12 13
14
Conflict of interest There is no conflict of interest to declare.
References 1 Gifford RW. Treatment of hypertensive emergencies, including use of sodium nitroprusside. Mayo Clin Proc 1959; 34: 387–394. 2 Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo Jr JL et al. The seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure: the JNC 7 report. JAMA 2003; 289: 2560–2572. 3 Cherney D, Straus S. Management of patients with hypertensive urgencies and emergencies: a systematic review of the literature. J Gen Intern Med 2002; 17: 937–945. 4 BASC. Interventions for deliberately altering blood pressure in acute stroke. Blood pressure in Acute Stroke Collaboration (BASC) (Cochrane Review). The Cochrane Library, Issue 3, 2001, CD000039, Update Software, John Wiley & Sons, Ltd. 5 Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M. ACC/AHA guidelines for the management of patients with ST-elevation myocardial
15
16
17
18 19
20
infarction—executive summary: a report of the American College of Cardiology/American Heart Association task force on practice guidelines (writing committee to revise the 1999 guidelines for the management of patients with acute myocardial infarction). Circulation 2004; 110: 588–636. Dalen JE, Howe III JP. Dissection of the aorta. Current diagnostic and therapeutic approaches. JAMA 1979; 242: 1530–1532. Mattu A, Martinez JP, Kelly BS. Modern management of cardiogenic pulmonary edema. Emerg Med Clin North Am 2005; 23: 1105–1125. Wilson SR, Hirsch NP, Appleby I. Management of subarachnoid haemorrhage in a non-neurosurgical centre. Anaesthesia 2005; 60: 470–485. Angeli P, Chiesa M, Caregaro L, Merkel C, Sacerdoti D, Rondana M et al. Comparison of sublingual captopril and nifedipine in immediate treatment of hypertensive emergencies. A randomized, single-blind clinical trial. Arch Intern Med 1991; 151: 678–682. Beltrame JF, Zeitz CJ, Unger SA, Brennan RJ, Hunt A, Moran JL et al. Nitrate therapy is an alternative to furosemide/morphine therapy in the management of acute cardiogenic pulmonary edema. J Card Fail 1998; 4: 271–279. Elliott WJ, Weber RR, Nelson KS, Oliner CM, Fumo MT, Gretler DD et al. Renal and hemodynamic effects of intravenous fenoldopam versus nitroprusside in severe hypertension. Circulation 1990; 81: 970–977. Hamilton RJ, Carter WA, Gallagher EJ. Rapid improvement of acute pulmonary edema with sublingual captopril. Acad Emerg Med 1996; 3: 205–212. Hirschl MM, Binder M, Bur A, Herkner H, Mu¨llner M, Woisetschla¨ger C et al. Safety and efficacy of urapidil and sodium nitroprusside in the treatment of hypertensive emergencies. Intensive Care Med 1997; 23: 885–888. Hirschl MM, Schreiber W, Woisetschla¨ger C, Kaff A, Raab H. [Sublingual nitroglycerin or intravenous enalaprilat in preclinical treatment of hypertensive patients with pulmonary edema]. [German] Z Kardiol 1999; 88: 208–214. Marigliano V, Santilli D, Fiorani M, Ariani A, Cacciafesta M, Ferri C et al. Hypertensive emergencies in old age: effects of angiotensin converting enzyme inhibition. J Hypertens Suppl 1988; 6: S91–S93. McNair A, Krogsgaard AR, Hilden T, Nielsen PE. Severe hypertension with cerebral symptoms treated with furosemide, fractionated diazoxide or dihydralazine. Danish Multicenter Study. Acta Med Scand 1986; 220: 15–23. Nelson GI, Silke B, Ahuja RC, Hussain M, Taylor SH. Haemodynamic advantages of isosorbide dinitrate over frusemide in acute heart-failure following myocardial infarction. Lancet 1983; 1: 730–733. Pastorelli R, Ferri C, Santucci A, Balsano F. New therapeutic possibilities in hypertensive emergencies. Curr Ther Res, Clin Exp 1991; 50: 857–868. Rubio-Guerra AF, Vargas-Ayala G, Lozano-Nuevo JJ, Narvaez-Rivera JL, Rodriguez-Lopez L. Comparison between isosorbide dinitrate aerosol and nifedipine in the treatment of hypertensive emergencies. J Hum Hypertens 1999; 13: 473–476. Schreiber W, Woisetschla¨ger C, Binder M, Kaff A, Raab H, Hirschl MM. The nitura study—effect of nitroglycerin or urapidil on hemodynamic, metabolic and respiratory parameters in hypertensive patients with Journal of Human Hypertension
Different interventions for hypertensive emergencies MI Perez and VM Musini 606
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
pulmonary edema. Intensive Care Med 1998; 24: 557–563. Verma SP, Silke B, Hussain M. First-line treatment of left ventricular failure complicating acute myocardial infarction: a randomised evaluation of immediate effects of diuretic, venodilator, arteriodilator, and positive inotropic drugs on left ventricular function. J Cardiovasc Pharmacol 1987; 10: 38–46. Wu SG, Lin SL, Shiao WY, Huang HW, Lin CF, Yang YH. Comparison of sublingual captopril, nifedipine and prazosin in hypertensive emergencies during hemodialysis. Nephron 1993; 65: 284–287. Yang HJ, Kim JG, Lim YS, Ryoo E, Hyun SY, Lee G. Nicardipine versus nitroprusside infusion as antihypertensive therapy in hypertensive emergencies. J Int Med Res 2004; 32: 118–123. Bussmann WD, Kenedi P, von Mengden HJ, Nast HP, Rachor N. Comparison of nitroglycerin with nifedipine in patients with hypertensive crisis or severe hypertension. Clin Investig 1992; 70: 1085–1088. Conen D, Ru¨ttimann S, Noll G, Schneider K, Mu¨ller J. Short- and long-term cerebrovascular effects of nitrendipine in hypertensive patients. J Cardiovasc Pharmacol 1988; 12: S64–S68. Dadkar VN, Karnik ND, Izar M, Sharma SR, Gandhi YP, Narawane NM et al. Sublingual nifedipine and captopril in hypertensive urgencies and emergencies. Indian Heart J 1993; 45: 185–187. Marghli S, Bouida W, Elatrous S, Boujdaria R, Nouira S, Abroug F. A prospective, randomized and controlled study of three intravenous antihypertensive drugs in severe hypertension. Reanimation Urgences 1997; 6: 289–296. Moritz RD, de Queiroz LP, Pereira MR, Scotinni MA. [Comparative study of the use of nifedipine and captopril in hypertensive emergencies]. [Portuguese] Arq Bras Cardiol 1989; 52: 323–326. Neutel JM, Smith DH, Wallin D, Cook E, Ram CV, Fletcher E et al. A comparison of intravenous nicardipine and sodium nitroprusside in the immediate treatment of severe hypertension. Am J Hypertens 1994; 7: 623–628. Nielsen PE, Krogsgaard A, McNair A, Hilden T. Emergency treatment of severe hypertension evaluated in a randomized study. Acta Med Scand 1980; 208: 473–480. Panacek EA, Bednarczyk EM, Dunbar LM, Foulke GE, Holcslaw TL. Randomized, prospective trial of fenoldopam vs sodium nitroprusside in the treatment of acute severe hypertension. Fenoldopam Study Group. Acad Emerg Med 1995; 2: 959–965. Perez C, Dougnac A, Alvarez M, Andresen M, Dı´az O, Geni R et al. [Sublingual captopril versus nifedipine in the treatment of hypertensive crisis]. [Spanish] Rev Med Chil 1991; 119: 402–405. Risler T, Bohm R, Wetzchewald D, Nast HP, Koch HH, Stein G et al. A comparison of the antihypertensive efficacy and safety of felodipine IV and nifedipine IV in patients with hypertensive crisis or emergency not responding to oral nifedipine. Eur J Clin Pharmacol 1998; 54: 295–298. Rohr G, Reimnitz P, Blanke P. Treatment of hypertensive emergency. Comparison of a new dosage form of the calcium antagonist nitrendipine with nifedipine capsules. Intensive Care Med 1994; 20: 268–271. Spah F, Grosser KD. Treatment of hypertensive urgencies and emergencies with nitrendipine, nifedi-
Journal of Human Hypertension
36
37
38
39
40
41
42
43 44 45
46
47
48
49
pine, and clonidine: effect on blood pressure and heart rate. J Cardiovasc Pharmacol 1988; 12: S154–S156. Ceyhan B, Karaaslan Y, Caymaz O, Oto A, Oram E, Oram A et al. Comparison of sublingual captopril and sublingual nifedipine in hypertensive emergencies. Jpn J Pharmacol 1990; 52: 189–193. Guerrera G, Melina D, Capaldi L, Mauro R, Colivicchi F, Cardillo C et al. [Sublingually administered captopril versus nifedipine in hypertension emergencies]. [Italian] Minerva Cardioangiol 1990; 38: 37–44. Pascale C, Zampaglione B, Marchisio M. Management of hypertensive crisis: nifedipine in comparison with captopril, clonidine and furosemide. Curr Ther Res, Clin Exp 1992; 51: 9–18. Pilmer BL, Green JA, Panacek EA, Elliot WJ, Murphy MB, Rutherford W et al. Fenoldopam mesylate versus sodium nitroprusside in the acute management of severe systemic hypertension. J Clin Pharmacol 1993; 33: 549–553. Pujadas R, Jane´ J, Forno´s C, Gago MJ, de la Concepcio´n N. Comparison of sublingual captopril and nifedipine in hypertensive crises. Arch Intern Med 1987; 147: 175–176. Reisin E, Huth MM, Nguyen BP, Weed SG, Gonzalez FM. Intravenous fenoldopam versus sodium nitroprusside in patients with severe hypertension. Hypertension 1990; 15: I59–I62. Zampaglione B, Pascale C, Marchisio M, Santoro A. The use of lacidipine in the management of hypertensive crises: a comparative study with nifedipine. J Cardiovasc Pharmacol 1994; 23: S116–S118. Franklin C, Nightingale S, Mamdani B. A randomized comparison of nifedipine and sodium nitroprusside in severe hypertension. Chest 1986; 90: 500–503. Bertel O, Conen D, Radu¨ EW, Mu¨ller J, Lang C, Dubach UC. Nifedipine in hypertensive emergencies. Br Med J (Clin Res Ed) 1983; 286: 19–21. Borghi C, Bacchelli S, Esposti DD, Bignamini A, Magnani B, Ambrosioni E. Effects of the administration of an angiotensin-converting enzyme inhibitor during the acute phase of myocardial infarction in patients with arterial hypertension. SMILE Study Investigators. Survival of Myocardial Infarction Long-term Evaluation. Am J Hypertens 1999; 12: 665–672. Lisk DR, Grotta JC, Lamki LM, Tran HD, Taylor JW, Molony DA et al. Should hypertension be treated after acute stroke? A randomized controlled trial using single photon emission computed tomography. Arch Neurol 1993; 50: 855–862. Nelson GIC, Nerma SP, Silke B, Abdulan S, Taylor SH. Management of hypertension complicating acute myocardial infarction (MI): comparison of sympathetic (labetalol) and calcium channel (nifedipine) blockade. [abstract] Aust NZ J Med (Suppl) 1984; 14: 577. Cotter G, Metzkor E, Kaluski E, Faigenberg Z, Miller R, Simovitz A et al. Randomised trial of high-dose isosorbide dinitrate plus low-dose furosemide versus high-dose furosemide plus low-dose isosorbide dinitrate in severe pulmonary oedema. Lancet 1998; 351: 389–393. Yoshida N. Antihypertensive effect of the long-acting Ca antagonist amlodipine vs sustained-release nifedipine in patients with acute aortic dissection. Rinsho to Kenkyu (Jpn J Clin Exp Med) 1998; 75: 1419–1422.
Different interventions for hypertensive emergencies MI Perez and VM Musini 607
50 Annane D, Bellissant E, Pussard E, Asmar R, Lacombe F, Lanata E et al. Placebo-controlled, randomized, double-blind study of intravenous enalaprilat efficacy and safety in acute cardiogenic pulmonary edema. Circulation 1996; 94: 1316–1324. 51 Krogsgaard AR, Hilden T, McNair A, Nielsen PE. Cerebral symptoms and blood pressure during parenteral administration of chlorpromazine, dihydralazine and diazoxide. Danish Mmulticentre Study on acute severe hypertension. Acta Med Scand 1983; 678(Suppl): 51–60. 52 Krogsgaard AR, McNair A, Hilden T, Nielsen PE. Reversibility of cerebral symptoms in severe hypertension in relation to acute antihypertensive therapy. Danish Multicenter Study. Acta Med Scand 1986; 220: 25–31.
53 McNair A, Krogsgaard AR, Hilden T, Nielsen PE. Reversibility of cerebral symptoms in severe hypertension in relation to acute antihypertensive therapy. Danish Multicenter Study. Acta Medica Scandinavica Supplementum 1985; 693: 107–110. 54 Duley L, Henderson-Smart DJ, Meher S. Drugs for treatment of very high blood pressure during pregnancy (Cochrane Review). The Cochrane Library, Issue 3, 2006, CD001449, Update Software, John Wiley & Sons, Ltd. 55 Roberts I, Yates D, Sandercock P, Farrell B, Wasserberg J, Lomas G et al. Effect of intravenous corticosteroids on death within 14 days in 10 008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet 2004; 364: 1321–1328.
Journal of Human Hypertension